Short-range wireless proximity networks typically involve devices that have a communications range of one hundred meters or less. To provide communications over long distances, these proximity networks often interface with other networks. For example, short-range networks may interface with cellular networks, wireline telecommunications networks, and the Internet.
IEEE 802.15.3 defines an ad hoc wireless short-range network (referred to as a piconet) in which a plurality of devices may communicate with each other. The timing of piconets is based on a repeating pattern of “superframes” in which the network devices may be allocated communications resources. Currently, the MultiBand OFDM Alliance (MBOA) is defining a media access control (MAC) layer for Ultra Wide Band (UWB) radios in relation with the IEEE 802.15.3. More information about Multiband OFDM can be found from http://www.multibandofdm.org/.
The first version of the MAC specifies a superframe that is 65536 microseconds in duration. This superframe contains 256 equally spaced Media Access Slots (MAS). Each MAS can be used for data communication. Thus, the length of each MAS is 256 microseconds.
According to this initial MAC version, the first eight MASs of the superframe are always reserved for the transmission of beacons. The time period covering these eight slots is called a Beacon Period (BP). During the Beacon Period, each MAS contains three beacon slots. Therefore, the total number of available beacon slots per superframe is twenty four. During a beacon slot, only a single beacon can be sent.
Unfortunately, because the initial MAC is inflexible in the number of beacon slots, it is likely that for any given situation, there will be either too many or too few available beacon slots. For instance, the maximum number of twenty four beacon slots specified by the initial MAC version is for devices that are located within two hops of each other. This number of beacon slots may be sufficient because the operating range provided by the UWB physical layer (PHY) is small. However, a greater number of beacon slots may sometimes be needed. This could be the case in scenarios involving, for example, rush hour trains or buses.
Conversely, situations may occur where there are too many beacon slots. For instance, when a network includes just one device, only two beacon slots are necessary during the BP. One of these slots is for beaconing and the other is for another device's beacon when it joins the network. Similarly, if two devices are present in a network, then only two slots during the BP are in use. Additionally, one slot must be reserved for a new device.
The initial MAC version specifies that every device operating in an active mode must listen to every beacon slot in a BP. Therefore, in the above mentioned situation involving two devices, the initial MAC version requires the two devices to monitor (or “listen to”) twenty two empty beacon slots. This results in unnecessary power consumption for the two devices.
Accordingly, the current inflexibility in the number of beacon slots can have the unfortunate effect of increasing device power consumption and/or decreasing available communications capacity. These effects are highly likely given the general nature of ad hoc networks, in which situations vary greatly.